Draft: Add scPortrait feature for issue #26

conf/modules.config

@@ -267,6 +267,15 @@ process {
         ext.args   = { "--diameter 9 --channel_axis 0 --save_flows" }
     }
 
+    // ---------------------------- scportrait -------------------------------------
+
+    withName: SCPORTRAIT_SEGMENT {
+        publishDir = [
+            path: { "${params.outdir}/${params.mode}/scportrait" },
+            mode: params.publish_dir_mode,
+        ]
+    }
+
     // ---------------------------- opt -----------------------------------------
 
     withName: OPT_FLIP {

modules/local/scportrait/imagesegment/main.nf

@@ -0,0 +1,36 @@
+process SCPORTRAIT_IMAGESEGMENT {
+    tag "$meta.id"
+    label 'process_high'
+    maxForks params.restrict_concurrency ? 1 : 0
+
+    container "docker.io/library/python:3.11-slim"
+
+    input:
+    tuple val(meta), path(image)
+
+    output:
+    tuple val(meta), path("${prefix}/nuclei_labels.tif"), emit: nuclei, optional: true
+    tuple val(meta), path("${prefix}/cells_labels.tif"), emit: cells, optional: true
+    path "versions.yml", emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    prefix = task.ext.prefix ?: "${meta.id}"
+    def nucleusOnly = params.nucleus_segmentation_only
+    template('image_segmentation.py')
+
+    stub:
+    prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    mkdir -p ${prefix}
+    touch ${prefix}/nuclei_labels.tif
+    touch ${prefix}/cells_labels.tif
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        scportrait: stub
+    END_VERSIONS
+    """
+}

modules/local/scportrait/imagesegment/meta.yml

@@ -0,0 +1,77 @@
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/meta-schema.json
+# # TODO nf-core: Add a description of the module and list keywords
+name: "scportrait_imagesegment"
+description: write your description here
+keywords:
+  - sort
+  - example
+  - genomics
+tools:
+  ## TODO nf-core: Add a description and other details for the software below
+  - "scportrait":
+      description: ""
+      homepage: ""
+      documentation: ""
+      tool_dev_url: ""
+      doi: ""
+      licence: null
+      identifier: null
+
+input:
+  ### TODO nf-core: Add a description of all of the variables used as input
+  - - meta:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. `[ id:'sample1' ]`
+    - bam:
+        type: file
+        description: Sorted BAM/CRAM/SAM file
+        pattern: "*.{bam,cram,sam}"
+        ontologies:
+          - edam: "http://edamontology.org/format_2572" # BAM
+          - edam: "http://edamontology.org/format_2573" # CRAM
+          - edam: "http://edamontology.org/format_3462" # SAM
+
+output:
+  ### TODO nf-core: Add a description of all of the variables used as output
+  bam:
+    - - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. `[ id:'sample1' ]`
+      - "*.bam":
+          type: file
+          description: Sorted BAM/CRAM/SAM file
+          pattern: "*.{bam,cram,sam}"
+          ontologies:
+            - edam: "http://edamontology.org/format_2572" # BAM
+            - edam: "http://edamontology.org/format_2573" # CRAM
+            - edam: "http://edamontology.org/format_3462" # SAM
+  versions_scportrait:
+    - - "${task.process}":
+          type: string
+          description: The name of the process
+      - "scportrait":
+          type: string
+          description: The name of the tool
+      - "scportrait --version":
+          type: eval
+          description: The expression to obtain the version of the tool
+
+topics:
+  versions:
+    - - ${task.process}:
+          type: string
+          description: The name of the process
+      - scportrait:
+          type: string
+          description: The name of the tool
+      - scportrait --version:
+          type: eval
+          description: The expression to obtain the version of the tool
+authors:
+  - "@TPbiocode"
+maintainers:
+  - "@TPbiocode"

modules/local/scportrait/imagesegment/templates/image_segmentation.py

@@ -0,0 +1,149 @@
+#!/usr/bin/env python3
+
+from pathlib import Path
+from importlib.metadata import PackageNotFoundError, version
+
+import numpy as np
+from tifffile import imread, imwrite
+
+from scportrait.pipeline.project import Project
+from scportrait.pipeline.segmentation.workflows import (
+	CytosolSegmentationCellpose,
+	DAPISegmentationCellpose,
+)
+
+
+def normalize_input_shape(image: np.ndarray) -> np.ndarray:
+	image = np.asarray(image)
+	image = np.squeeze(image)
+
+	if image.ndim == 2:
+		return image[np.newaxis, ...]
+
+	if image.ndim == 3:
+		if image.shape[0] <= 4:
+			return image
+		if image.shape[-1] <= 4:
+			return np.moveaxis(image, -1, 0)
+		raise ValueError(f"Unable to infer channel axis from image shape {image.shape}")
+
+	if image.ndim == 4:
+		image = image[0]
+		if image.shape[0] <= 4:
+			return image
+		if image.shape[-1] <= 4:
+			return np.moveaxis(image, -1, 0)
+		raise ValueError(f"Unable to infer channel axis from 4D image shape {image.shape}")
+
+	raise ValueError(f"Unsupported morphology image shape {image.shape}")
+
+
+def extract_label(label_obj) -> np.ndarray:
+	if hasattr(label_obj, "scale0"):
+		label_obj = label_obj.scale0.image
+	elif hasattr(label_obj, "image"):
+		label_obj = label_obj.image
+
+	data = getattr(label_obj, "data", label_obj)
+	if hasattr(data, "compute"):
+		data = data.compute()
+
+	data = np.asarray(data)
+	data = np.squeeze(data)
+	return data.astype(np.uint32)
+
+
+def run_segmentation(
+	image_path: str,
+	nucleus_only: bool,
+	nuclei_out: str,
+	cells_out: str,
+	project_dir: str,
+) -> None:
+	image = normalize_input_shape(imread(image_path))
+
+	if nucleus_only:
+		segmentation_cls = DAPISegmentationCellpose
+		config = {
+			"DAPISegmentationCellpose": {
+				"input_channels": 1,
+				"output_masks": 1,
+				"cache": ".",
+				"chunk_size": 100,
+				"nucleus_segmentation": {"model": "nuclei"},
+			}
+		}
+		image = image[:1, :, :]
+		channel_names = ["nucleus"]
+	else:
+		segmentation_cls = CytosolSegmentationCellpose
+		if image.shape[0] == 1:
+			image = np.repeat(image, 2, axis=0)
+		else:
+			image = image[:2, :, :]
+
+		config = {
+			"CytosolSegmentationCellpose": {
+				"input_channels": 2,
+				"output_masks": 2,
+				"cache": ".",
+				"chunk_size": 100,
+				"nucleus_segmentation": {"model": "nuclei"},
+				"cytosol_segmentation": {"model": "cyto2"},
+				"match_masks": True,
+				"filter_masks_size": False,
+			}
+		}
+		channel_names = ["nucleus", "cytosol"]
+
+	project = Project(
+		project_location=project_dir,
+		config_path=config,
+		segmentation_f=segmentation_cls,
+		overwrite=True,
+		debug=False,
+	)
+	project.load_input_from_array(image, channel_names=channel_names)
+	project.segment()
+
+	if "seg_all_nucleus" in project.sdata:
+		imwrite(nuclei_out, extract_label(project.sdata["seg_all_nucleus"]))
+
+	if "seg_all_cytosol" in project.sdata:
+		imwrite(cells_out, extract_label(project.sdata["seg_all_cytosol"]))
+
+
+def generate_versions_yml() -> None:
+	try:
+		scportrait_version = version("scportrait")
+	except PackageNotFoundError:
+		scportrait_version = "unknown"
+
+	with open("versions.yml", "w", encoding="utf-8") as handle:
+		handle.write('"${task.process}":\n')
+		handle.write(f"    scportrait: {scportrait_version}\n")
+
+
+def main() -> None:
+	image_path = "${image}"
+	prefix = "${prefix}"
+	project_dir = "scportrait_project"
+	nucleus_only = "${nucleusOnly}".lower() == "true"
+	nuclei_out = f"{prefix}/nuclei_labels.tif"
+	cells_out = f"{prefix}/cells_labels.tif"
+
+	Path(prefix).mkdir(parents=True, exist_ok=True)
+	Path(project_dir).mkdir(parents=True, exist_ok=True)
+
+	run_segmentation(
+		image_path=image_path,
+		nucleus_only=nucleus_only,
+		nuclei_out=nuclei_out,
+		cells_out=cells_out,
+		project_dir=project_dir,
+	)
+	generate_versions_yml()
+
+
+if __name__ == "__main__":
+	main()

modules/local/scportrait/imagesegment/tests/main.nf.test

@@ -0,0 +1,79 @@
+// TODO nf-core: Once you have added the required tests, please run the following command to build this file:
+// nf-core modules test scportrait/imagesegment
+nextflow_process {
+
+    name "Test Process SCPORTRAIT_IMAGESEGMENT"
+    script "../main.nf"
+    process "SCPORTRAIT_IMAGESEGMENT"
+
+    tag "modules"
+    tag "modules_"
+    tag "scportrait"
+    tag "scportrait/imagesegment"
+
+    // TODO nf-core: Change the test name preferably indicating the test-data and file-format used
+    test("sarscov2 - bam") {
+
+        // TODO nf-core: If you are created a test for a chained module
+        // (the module requires running more than one process to generate the required output)
+        // add the 'setup' method here.
+        // You can find more information about how to use a 'setup' method in the docs (https://nf-co.re/docs/contributing/modules#steps-for-creating-nf-test-for-chained-modules).
+
+        when {
+            process {
+                """
+                // TODO nf-core: define inputs of the process here. Example:
+
+                input[0] = [
+                    [ id:'test' ],
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/bam/test.paired_end.sorted.bam', checkIfExists: true),
+                ]
+                """
+            }
+        }
+
+        then {
+            assert process.success
+            assertAll(
+                { assert snapshot(
+                    process.out,
+                    path(process.out.versions[0]).yaml
+                ).match() }
+                //TODO nf-core: Add all required assertions to verify the test output.
+                // See https://nf-co.re/docs/contributing/tutorials/nf-test_assertions for more information and examples.
+            )
+        }
+
+    }
+
+    // TODO nf-core: Change the test name preferably indicating the test-data and file-format used but keep the " - stub" suffix.
+    test("sarscov2 - bam - stub") {
+
+        options "-stub"
+
+        when {
+            process {
+                """
+                // TODO nf-core: define inputs of the process here. Example:
+
+                input[0] = [
+                    [ id:'test' ],
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/bam/test.paired_end.sorted.bam', checkIfExists: true),
+                ]
+                """
+            }
+        }
+
+        then {
+            assert process.success
+            assertAll(
+                { assert snapshot(
+                    process.out,
+                    path(process.out.versions[0]).yaml
+                ).match() }
+            )
+        }
+
+    }
+
+}

nextflow.config

@@ -45,7 +45,7 @@ params {
     format                     = 'xenium'   // preset value set as `xenium`
 
     // Segmentation methods
-    image_seg_methods          = ["cellpose", "xeniumranger", "baysor"]
+    image_seg_methods          = ["cellpose", "xeniumranger", "baysor", "scportrait"]
     transcript_seg_methods     = ["proseg", "segger", "baysor"]
     segfree_methods            = ["ficture", "baysor"]
 

nextflow_schema.json

@@ -36,7 +36,7 @@
                 },
                 "method": {
                     "type": "string",
-                    "enum": ["cellpose", "xeniumranger", "baysor", "proseg", "segger", "ficture"],
+                    "enum": ["cellpose", "xeniumranger", "baysor", "scportrait", "proseg", "segger", "ficture"],
                     "description": "Segmentation method to run."
                 },
                 "gene_panel": {
@@ -174,7 +174,7 @@
                     "type": "array",
                     "items": {
                         "type": "string",
-                        "enum": ["cellpose", "xeniumranger", "baysor"]
+                        "enum": ["cellpose", "xeniumranger", "baysor", "scportrait"]
                     },
                     "description": "List of image-based segmentation methods."
                 },